Comparative Study of Three Different Methods for the Determination

A fume hood should be used, and gloves, protective clothing, and eyewear should be ...... Aflatoxins contamination in Pakistani brown rice: a comparis...
0 downloads 0 Views 55KB Size
J. Agric. Food Chem. 2002, 50, 3375−3379

3375

Comparative Study of Three Different Methods for the Determination of Aflatoxins in Tahini DILARA NILU¨ FER

AND

DILEK BOYACıOGˇ LU*

Department of Food Engineering, Istanbul Technical University, 80626 Maslak, Istanbul, Turkey

Aflatoxins spiked at three different levels (6.5, 13.0, and 19.5 µg/kg) in tahini, a sesame butter, were analyzed by using three different methods: high-performance liquid chromatography (HPLC), fluorometry, and enzyme-linked immunosorbent assay (ELISA). An immunoaffinity column was used for cleanup and purification of extracts prior to detection by HPLC and fluorometry. All methods were statistically evaluated for accuracy, precision, and simple correlations. Additionally, 14 tahini samples randomly obtained from Turkish retail markets were analyzed using an immunoaffinity column cleanup procedure coupled with the HPLC detection method. The fluorometric determination method involving an immunoaffinity column cleanup step was found to be highly correlated with the HPLC method (r ) 0.978). Both methods were found to be effective due to their high recoveries and low variance for the prediction of total aflatoxin contamination in tahini samples. The ELISA method, due to its high variation in replicates, was found to be applicable only as a screening method. The survey study demonstrated the need for control of aflatoxin contamination of foodstuffs involving sesame seeds as an ingredient. KEYWORDS: Aflatoxin; analysis; sesame seeds; tahini; high-performance liquid chromatography (HPLC); fluorometry; enzyme-linked immunosorbent assay (ELISA); immunoaffinity column

INTRODUCTION

Aflatoxins B1, G1, B2, and G2, which are the secondary metabolites of the molds Aspergillus flaVus, Aspergillus parasiticus, and Aspergillus nomius, produce great risk by contaminating a wide variety of agricultural commodities and foodstuffs, especially those having high carbohydrate and/or fat contents, such as tree nuts and nut products, peanuts, corn, cereals, grains, oilseeds, dried figs and raisins, cottonseed, milk, feedstuffs, and dried spices (1-8). Tahini (sesame butter) produced by milling of dehulled and roasted sesame seeds is widely consumed as a mixture with sweet foods such as honey and grape molasses and also is an important ingredient of halva, which is a popular sweet food in Turkey and in some Middle East countries (9, 10). A. flaVus and A. parasiticus have been reported to grow on sesame seed and produce aflatoxins (11-13). Several survey studies on sesame seeds (14-17) and tahini (18) have shown the extent of aflatoxin contamination. In recent years, there has been an increase in demand for simple, quick, accurate, and specific methods for the determination of aflatoxins. The immunological methods that have been used since the 1990s are accepted as official methods for aflatoxin determination in some food commodities; however, it is still necessary to evaluate their efficiency in different foodstuffs and to correlate their results with other accepted analytical methods (19). High-performance liquid chromatog* Corresponding author (telephone 90 212 285 6039; fax 90 212 285 2925; e-mail [email protected]).

raphy (HPLC) and enzyme-linked immunosorbent assay (ELISA) methods were found to be highly correlated for the analysis of total aflatoxins in foodstuffs such as peanut, peanut butter, and corn and for aflatoxin B1 in corn and mixed feeds, whereas in other foodstuffs, such as cereals, low correlations were observed (20). ELISA methods could be used only for screening purposes with the detection limit above the regulatory limits because they are not reliable when used as a quantitative method (21, 22). Analyses of aflatoxins B1, B2, and G1 in corn and of total aflatoxin levels in peanut butter and aflatoxin B1 in corn and roasted peanuts based on ELISA methods were adopted as AOAC Official Methods (23). Immunoaffinity column cleanup techniques coupled with HPLC detection or fluorometric detection are applied to various food matrices such as corn, peanuts, peanut butter, pistachios, Brazil nuts, almonds, nut confectionery products, maize gluten, soya expeller, beer, animal products, animal feedstuffs, and dried fruits such as figs, dates, and apricots (22, 24-26). The immunoaffinity column (Aflatest) procedure is an AOAC-approved method for the determination of aflatoxins in corn, raw peanuts, and peanut butter (23). The objective of this study was to analyze aflatoxins in tahini, which is a traditional Turkish food, by HPLC, fluorometry, and ELISA methods. Correlations of three aflatoxin determination methods were investigated, and methods were compared in terms of accuracy and precision. A limited aflatoxin survey of tahini in 14 samples randomly collected from the market was also carried out.

10.1021/jf020005a CCC: $22.00 © 2002 American Chemical Society Published on Web 05/02/2002

3376

J. Agric. Food Chem., Vol. 50, No. 12, 2002

Nilu¨fer and Boyacıogˇlu

Table 1. Amount of Total Aflatoxins Added to the Samples and

Aflatoxin B1, B2, G1, and G2 Concentrations after Spiking level aflatoxin no. (ppb) 1 2 3

6.5 13.0 19.5

amount of total aflatoxins spiked to 25 g of sample (ng)

volume spiked from stock solution (mL/ 25 g of sample)

162.5 325 487.5

1.25 2.5 3.75

individual aflatoxin concn (ppb) B1 B2 G1 G2 2.5 5.0 7.5

0.75 1.5 2.25

2.5 5.0 7.5

0.75 1.5 2.25

MATERIALS AND METHODS Apparatus. For HPLC analysis an HPLC system equipped with a Waters 501 solvent delivery system, a 7010 injection port, an HP 1046A fluorescence detector (excitation, 363 nm; emission, 440 nm), a 300 mm × 3.9 mm i.d. Millipore-Waters µ-Bondapak C18 LC column, and 810 Baseline software was used. A Vicam series 4 fluorometer, model Vicam V1.0 (Vicam, Watertown, MA), was used for the fluorometric determination of total aflatoxins. As an ELISA microstrip reader, a Statfax 303 Plus (Awareness Technology Inc., Palm City, FL) with a 650 nm filter was utilized. Kits. Aflatest P columns (Vicam) and Veratox kits for aflatoxin (Neogen, Lansing, MI) were tested. Samples. Tahini samples of 14 different trademarks were randomly selected from the Turkish retail market. One sample confirmed to contain a total aflatoxin level below the detection limit was used for the comparison of methods. Due to the separation of oil-water phases during storage, samples were homogenized by mixing nearly for 30 min (Janke & Kunkel RW 20) prior to analysis. Twenty-five gram portions of samples were weighed into tightly capped small glass jars and wrapped with aluminum foil. Aflatoxin Standard Solution. An aflatoxin B and G mixture purchased in dry concentrate form (Sigma, St. Louis, MO) was dissolved in 7.5 mL of benzene/acetonitrile (98:2, v/v). Five hundred microliters of a mixed aflatoxins stock standard solution, containing 5 µg/mL aflatoxin B1, 1.5 µg/mL aflatoxin B2, 5 µg/mL aflatoxin G1, and 1.5 µg/mL aflatoxin G2 (B1/B2/G1/G2 10:3:10:3), was transferred into an amber vial and evaporated just to dryness under a stream of nitrogen at room temperature. The total aflatoxin concentration was 130 ng/mL when the contents of the vial were dissolved in 50 mL of benzene/acetonitrile (98:2, v/v). Spiking. Test samples were spiked with total aflatoxins B1, B2, G1, and G2 at 6.5, 13.0, and 19.5 µg/kg levels in triplicate. These spike levels represent values above and below the maximum allowable regulatory limits of the Turkish Food Codex, which is 10 µg/kg (27). Each 25 ( 0.1 g of sample was spiked with the known amounts of 130 ppb stock aflatoxin solution given in Table 1. The spiked samples were kept in a fume hood to allow the benzene/acetonitrile solvent to evaporate and were kept at 4 °C until analysis. Extraction. Aflatoxins were extracted from samples using AOAC Method 968.22 (28). Spiked samples were transferred from glass jars to the blender by rinsing the jar with 125 mL of methanol/water (70: 30, v/v). Five grams of NaCl was added and blended at high speed (23000 rpm) for 2 min. The sample extract was divided into three portions to be analyzed by three determination methods. Immunoaffinity Column Cleanup. Affinity column cleanup was performed according to AOAC Method 968.22 (28). Filtered and diluted extracts were passed through the affinity columns followed by HPLC determination and fluorometric determination methods. The methanol eluate was collected in 2 mL Eppendorf tubes and evaporated under a stream of nitrogen gas. The dry residue was kept at 4 °C until analysis. Derivatization. Aflatoxins were derivatized with trifluoroacetic acid (TFA) according to AOAC Method 990.33 with slight modifications prior to HPLC injections (23). The modification involved an increase in mixing time to 1 min and the use of 0.95 mL of a water/acetonitrile (9:1, v/v) solvent system, instead of 1.95 mL, to dissolve aflatoxins. HPLC Determination. For liquid chromatographic determination a water/acetonitrile/methanol (70:17:17, v/v) mobile phase system

having a flow rate of 1 mL/min was used (23). The retention times of each aflatoxin component (G1, B1, G2, and B2) were 6.3, 7.5, 9.9, and 12.7 min, respectively, in a 20 min run time. Toxins were detected by fluorescence at an excitation wavelength of 363 nm and emission wavelength of 440 nm. The injection volume was 50 µL. Each sample was injected twice. Fluorometric Determination. The dry sample was dissolved in 1 mL of HPLC grade methanol and mixed in a minishaker for 1 min at 1600 rpm. One milliliter of developer solution was added to the sample extract and mixed at 1600 rpm for 30 s. Following transfer of sample extract and developer solution into the fluorometry cuvette, the total aflatoxin content of the sample was read out in parts per billion. Each sample was measured twice (29). ELISA Determination. ELISA analysis was performed according to the instructions of the Neogen Veratox aflatoxin procedure. Kits and extracts were brought to ambient temperature before analysis. A multichannel pipettor was used for ELISA analysis, and great attention was given to the incubation periods. Concentration of total aflatoxins in parts per billion was recorded from a 650 nm filter ELISA reader that was calibrated using aflatoxin standards (30). Survey Study. Fourteen tahini samples randomly collected from various markets in Istanbul were analyzed by immunoaffinity cleanup and HPLC determination using the same conditions previously described. Decontamination of Equipment. To avoid cross-contamination of samples, all glassware was soaked in a 10% solution of household bleach containing 5.25% NaOCl for a day, which was followed by washing with a mild detergent and rinsing with pure water (31). The container of the Waring blender was rinsed with methanol after washing with a mild detergent. The syringe barrel and the microsyringes (injector) were rinsed thoroughly with methanol (29). Statistical Analysis. The performances of three different aflatoxin determination methods were compared by statistical analysis with an SPSS software program. At each spiking level, the recoveries (percent) and the coefficients of variation (percent) were calculated. The differences between the HPLC, fluorometry, and ELISA methods at total aflatoxin levels were analyzed by one-way ANOVA experimental design, where the variance component was accepted as the concentration of aflatoxins. Significant differences were analyzed by comparing the mean values using Duncan’s new multiple-range test. A simple linear regression analysis was performed between the results of HPLC, fluorometry, and ELISA methods. RESULTS AND DISCUSSION

The results obtained for the HPLC method for each aflatoxin component and total aflatoxin levels of tahini samples measured by HPLC, fluorometry, and ELISA are given in Table 2. The coefficents of variation (CV%) of the HPLC analysis of aflatoxins B1, G1, B2, and G2 spiked at three levels range between 0.98 and 32.12%, between 7.81 and 31.82%, between 3.96 and 20.21%, and between 7.90 and 37.92%, respectively, and for total aflatoxins between 3.17 and 20.77%. In general, repeatability, as defined by the CV, of HPLC analysis of tahini samples for aflatoxin B1 and total aflatoxins decreased as the spiking levels decreased. The analytical variations of aflatoxins B1, G1, B2, and G2 obtained through immunoaffinity cleanup followed by HPLC detection are presented in Table 2. For aflatoxins B1, G1, and G2, values of CV% were at their lowest at the highest spike level, except for aflatoxin B2. Although the spiking levels in this study are below the spiking levels of the studies conducted on oilseeds and peanut butter samples by Trucksess et al. (31), Carman et al. (32), and Patey et al. (33), using immunoaffinity column cleanup coupled with the HPLC detection method, the CV% for tahini samples analyzed with HPLC are nearly the same or even less. Recovery values were calculated to determine the accuracy of the methods. For HPLC determination of aflatoxins in tahini, recovery values were in the range of 54.8-72.9% for aflatoxin

Analysis of Aflatoxins in Tahini

J. Agric. Food Chem., Vol. 50, No. 12, 2002

3377

Table 2. HPLC, Fluorometry, and ELISA Results of Tahini Samples Spiked with Three Different Levels of Aflatoxins B1, B2, G1, and G2 (10:3:10:3)

along with Standard Deviations for Repeatability (Sr), Recovery Values (Percent), and Coefficients of Variation (CV%) HPLC resultsa aflatoxin (ppb) sample

fluorometry resultsa

ELISA resultsa

total aflatoxin (ppb)

total aflatoxin (ppb)

av

av

B1

B2

NDb ND

ND ND

ND ND

ND ND

ND ND

2.43 1.30 1.86 0.80

4.3 4.8 4.55 0.35

first level (6.5 ppb) B1, 2.5 ppb; B2, 0.75 ppb G1, 2.5 ppb; G2, 0.75 ppb mean Src recovery % CV%d

1.02 1.87 1.24 1.37 0.441 54.8 32.12

0.70 0.72 0.67 0.69 0.028 92.0 3.96

1.19 1.96 1.44 1.53 0.391 61.2 25.52

0.72 0.73 0.54 0.66 0.110 88.0 16.61

3.63 5.28 3.88 4.26 0.885 65.4 20.77

9.50 10.00 6.50 8.67 1.89 133.3 21.84

13.8 6.1 9.2 9.70 3.87 149.23 39.94

second level (13 ppb) B1, 5 ppb; B2, 1.5 ppb G1, 5 ppb; G2, 1.5 ppb mean Sr recovery % CV%

3.42 3.02 3.02 3.15 0.230 63.0 7.31

1.45 0.97 1.17 1.20 0.242 80.0 20.21

3.68 1.94 2.65 2.76 0.877 55.2 31.82

1.21 0.59 1.32 1.04 0.393 69.3 37.92

9.76 6.72 8.16 8.21 1.522 62.7 18.54

17.75 15.50 16.75 16.67 1.13 128.2 6.76

13.1 20.7 23.0 18.93 5.18 145.64 27.36

third level (19.5 ppb) B1, 7.5 ppb; B2, 2.25 ppb G1, 7.5 ppb; G2, 2.25 ppb mean Sr recovery % CV%

5.53 5.47 5.42 5.47 0.054 72.9 0.98

2.14 2.00 1.82 1.99 0.159 88.4 8.00

6.35 5.48 6.23 6.02 0.470 80.3 7.81

1.76 1.89 1.61 1.75 0.138 77.8 7.90

15.78 14.84 15.09 15.24 0.483 78.1 3.17

25.00 26.25 26.00 25.75 0.66 132.1 2.57

36.4 36.4 21.6 31.47 8.54 161.37 27.16

first control second control mean Sr

G1

G2

total

a The average of two injections or two readouts. b Not detected (detection limit for aflatoxin B ,